This is a book of children's literature. This appears to be volume 6 of a series. LC Control Number: 22011853
Publisher Chicago, The Bookhouse for Children
Possible copyright status NOT_IN_COPYRIGHT
Book contributor Internet Archive
Full catalog record MARCXML
This books is being used for tests in data processing and comparing results. The original scan was done on a beta release of the Scribe station that uses Canon 1ds MarkII cameras to produce cr2 RAW files. These are then processed into the jp2 files. There are now multiple PDF files that can be used to see the differences in different processing approaches.
In the reviews of this item, each file will be introduced so that there can be a place to comment on each approach.
Thank you all for helping build an open access library.
This book has an
editable web page
April 26, 2006
...help with explanation of previous review...
Hello - I am new to this site so please excuse the fact that I may be posting this question in the wrong place. The previous reviewer states, "I grew up being told to be very careful with this book, and 9 to 11 others like it." --- Is the reviewer referring to being careful of the content of the book (possible inappropriate passages for children) or being careful not to damage the book because it is high-quality in terms of the illustrations, etc? Thank you.
December 26, 2005
About the Series
I grew up being told to be very careful with this book, and 9 to 11 others like it. The illustrations are fascinating, and I have memories of many hours of reading these books.
My parents still have the entire series somewhere. It is a treasure.
June 11, 2005
DJVU version descriptions
There are currently 3 djvu options for downloading. This note is to describe each and to solicit responses on what should be done better or worse with them.
This one is called djvu (download):
Is produced by the lizardtech djvuencoder. It includes OCR output so it can be searched. It is served as a download so that it will fire up the lizardtech djvu browser plug-in. If you are unfamiliar with djvu, it is a (pretty) open format for representing scanned documents. it is very compressed and clever. there are open source encoders and decoders, and lizardtech makes proprietary versions. It is more tuned to scanned documents than PDF, but adobe is catching up.
The APPLET version of this:
is the same document, but offered in a way that fires up a java applet to view it. this is slower, but does not require a download and install. The javadjvu project was started by a gifted USF student and continued by bill riemers with lizardtech and Internet Archive support. It is an opensource sourceforge project, and we could really use help extending this. Bill will work on bugs and speed/functionality enhancements under an Archive contract through some of 2005. http://javadjvu.sourceforge.net
Another djvu version was submitted by Jim Rile, a djvu and pdf guru based on a different encoding system.
Please try out his version and make comments on whether you find it better, worse, or the same.
June 8, 2005
About the Book:
The Latch Key is a beautiful book about children's literature with colored illustrations. The first third is about the lives of authors such as Aesop (he was thrown to his death from a precipice by the Delphians), Hawthorne, Lewis Carroll, Cervantes, etc... The second third is about children's literature, fairy tales, myth, Mother Goose, epics, etc..., and the final third is index material.
About the Scan:
I scanned the book using the Internet Archive's Scribe machine. It uses a Java user interface that controls two Canon EOS Mark IIds cameras to photograph the book, nondestructively, under glass. The PHP driven image processing pipeline creates the animated gif, the DjVu book, the zip of JPEG2000 cropped/deskewed images, the tar of raw CR2 files, and the OCR text.
About the technology:
The PDFs where constructed in various ways (thanks to our partners!) using various tools. HP has made over a thousand PDFs for us. There is a very nice one (thanks Jim Rile) that is only 17MG (smaller than a single raw photo!) We'd love an open-source automated mechanism driven off of the scandata.xml that puts the OCRed text under the JP2s in a linear PDF.
The JPEG 2000 images in the zip file are about 1.5MB each. Photoshop needs a plugin to read them, but ACDSee handles them fine. We have the DPI settings wrong in this batch (should be 400dpi). We could use some help getting this working in Imagemagick.
The metadata (marc and dublin core) was fetched by Sebastian Hammer of Index Data, from the Library of Congress.
We use LizardTech's DjVu software for the OCR text and DjVu creation. Dave Coffin's DCRaw and Leptonica's skew detection are also in the mix.
All of the Archive's book scanning software is open source and available on Sourceforge in the scribesw project.
This is our first book to go on-line from the Scribe machine, and it is close to representational of the quality we are shooting for. Thanks to the countless folks who have helped to make this book accessible to all.